The Initial Molecular Interaction between Mouse Sperm and the Zona Pellucida Is a Complex Binding Event*

Prior to fertilization, mammalian sperm must first bind to the zona pellucida (ZP), a glycoprotein matrix surrounding the egg. Sperm specifically bind to ZP3, an 83-kDa glycoprotein which functions as both an adhesion molecule and as a secretagogue for acrosomal exo- cytosis (Litscher, E. S., and Wassarman, P. M. (1993) Trends Glycosci. Glycotechnol. 5, 369–388). We used acid solubilized, 125 I-labeled ZPs to quantify the initial binding event on mouse spermatozoa. Live sperm could not be used since solubilized ZPs rapidly initiated exocytosis. Instead, acrosome intact mouse sperm were briefly fixed in 1% glutaraldehyde for binding studies using a standard filtration assay. The fixed sperm are suitable for sperm-zona binding assays based on two experiments: 1) incubating either live or fixed sperm in low concentrations of 125 I-ZPs not sufficient to induce acrosomal exocytosis revealed no differences in binding up to 15 min and 2) solubilized, unlabeled ZPs competed for 125 I-ZPs with an K I of approximately 3.78 n M . Sperm- 125 I-ZP binding reached equilibrium with a (cid:116) 1 ⁄ 2 of (cid:59) 22 min at 37 °C. Affinity parameters were calculated using the well substantiated assumption that only ZP3 binds intact mouse sperm. The on-rate constant for associa- tion of 125 I-ZP binding to the mouse sperm surface was calculated to be 3.2 (cid:51) 10 6 M (cid:50) 1 min (cid:50) 1 . The saturation binding isotherm revealed that there radioactivity contribution

I-ZPs with an K I of approximately 3.78 nM. Sperm-125 I-ZP binding reached equilibrium with a 1 ⁄2 of ϳ22 min at 37°C. Affinity parameters were calculated using the well substantiated assumption that only ZP3 binds intact mouse sperm. The on-rate constant for association of 125 I-ZP binding to the mouse sperm surface was calculated to be 3.2 ؋ 10 6 M ؊1 min ؊1 . The saturation binding isotherm revealed that there are approximately 30,000 binding sites, ascribed to ZP3, with an EC 50 of 1.29 nM. Further analysis indicated that this binding is complex (Hill coefficient ‫؍‬ 1.72), suggesting involvement of multiple receptors on the sperm surface and/or multiple ligand moieties. High and low affinity ZP binding sites on the sperm surface were confirmed by dissociation experiments. 125 I-ZP dissociation was clearly biphasic, and kinetic off-rate constants of 0.161 min ؊1 and 0.0023 min ؊1 were calculated for the low and high affinity sites, respectively. Apparent affinities (K d values) of 50 nM for the low affinity and 0.72 nM for the high affinity interaction were calculated from the rate constants. These data demonstrate that the initial adhesion event between mouse sperm and the zona pellucida is a high affinity event which is sufficient to tether a sperm to the extracellular matrix prior to the induction of acrosomal exocytosis.
When mammalian sperm first encounter an egg, they bind to the zona pellucida (ZP), 1 an extracellular glycoprotein matrix secreted by the egg (Inoue and Wolf, 1975;Gwatkin and Williams, 1977). This binding event is a receptor-ligand-mediated interaction that serves not only to physically anchor the sperm to the egg, but also to initiate a signal transduction pathway culminating in acrosomal exocytosis (Bleil and Wassarman, 1980a, 1980bFlorman and Storey, 1982). In the mouse, acrosomal exocytosis occurs after sperm are bound to the surface of the zona pellucida (Saling et al., 1979;Storey, 1981, 1982) and is thought to be required for the penetration of this egg vestment. Thus, the initial steps in fertilization use a receptor mediated stimulus secretion coupling system to control exocytosis of the acrosomal vesicle.
Following the demonstration that sperm were acrosome intact during initial binding to the zona pellucida (Saling et al., 1979), it was soon convincingly demonstrated that the 83-kDa glycoprotein component of the zona pellucida, ZP3, is the ligand for sperm binding (Bleil and Wassarman, 1980a, 1980bFlorman and Wassarman, 1985). Surprisingly, the sperm surface receptor for ZP3 remains elusive, although numerous candidates have appeared in the literature (Bleil and Wassarman, 1990;Cheng et al., 1994;Leyton and Saling, 1989b;Leyton et al., 1992;Kalab et al., 1994;Miller et al., 1992). In addition, there has been little biochemical characterization of mammalian sperm receptor-ZP3 interactions. Sperm-zona binding is mediated by one or more of the O-linked carbohydrate moieties on ZP3 (Florman and Wassarman, 1985); however, receptor aggregation may be required to initiate signaling since glycopeptides of ZP3 will bind to sperm, but not initiate the acrosome reaction (Florman and Wassarman, 1985;Leyton and Saling, 1989a).
Unlike other receptor ligand systems, direct assessment of receptor ligand affinities, number of binding sites, cooperativity of binding, or other characteristics of sperm-ZP3 binding, has not been undertaken. This appears unusual since the initial event of sperm-egg binding is prerequisite to all subsequent steps completing fertilization. In mammalian systems, however, two intrinsic factors have presented obstacles to obtaining direct biochemical characterization of the receptor-ligand interactions. First, ligand availability has been limited by the difficulties inherent in obtaining large quantities of purified mammalian oocyte components, in this case, the zona pellucida. Second, the loss, or modification, of receptor sites on live sperm following acrosomal exocytosis creates complex interactions between the receptor and ligand populations, making experimental design and interpretation untenable.
The development of relatively straightforward mass isolation techniques (Tanphaichitr et al., 1993), which we have further modified for increased efficiency and yield, has ameliorated the problem of ligand availability. The loss of receptor sites upon exposure to the ligand appeared to be a more intractable problem since the effects that addition of inhibitors of acrosomal exocytosis might have on the receptor-ligand interactions are unknown. Following the lead of researchers char-acterizing receptor-ligand binding in other systems (Hulme and Birdsall, 1992;Schroder et al., 1990;van Maurik et al., 1985;von Zastrow and Kobilla, 1992;Zidovetzki et al., 1991), we have turned to the use of fixed cells for binding studies. We have developed a sensitive, quantitative binding assay to evaluate sperm-ZP binding. This report presents the first direct quantitation of sperm-ZP binding, calculates affinity parameters for ZP3 binding based on the assumption that only ZP3 binds to acrosome intact sperm (Bleil and Wassarman 1980a, 1980bFlorman et al., 1984;Florman and Wassarman, 1985;Mortillo and Wassarman, 1991), and demonstrates the presence of both high and low affinity binding sites on acrosome-intact sperm.

Materials
Mice (ICR strain) were purchased from Harlan Sprague-Dawley (San Diego, CA). Na 125 I was purchased from DuPont NEN. Resins for gel filtration were purchased from Bio-Rad. Protease inhibitors, DNase, hyaluronidase, BSA, casein, lactalbumin, and Percoll were purchased from Sigma. Glutaraldehyde was purchased from either Electron Microscopy Sciences (Fort Washington, PA) or from Sigma. Chloramine-T was purchased from Aldrich. Whatman GF/C glass fiber filters for binding assays and all other chemicals were purchased from Fisher.

Isolation of Mouse Zonae Pellucidae
Zonae pellucidae were collected from ovarian homogenates of 3-week-old female mice according to Tanphaichitr et. al. (1993) with the modifications outlined below. Ovaries were homogenized, as described, with the addition of 20 g/ml leupeptin and aprotinin, and 200 g/ml benzamidine. The homogenate was layered onto a 3-step Percoll gradient (3 ml 25%, 4 ml 10%, 2 ml 3%) and centrifuged at 200 ϫ g for 2 h at 4°C. The 10% step, which contained virtually all the ZPs, was collected, diluted with buffer (1:12), and centrifuged at high speed (16,000 ϫ g, 10 min) in a standard microcentrifuge (Brinkmann Instruments) to pellet the ZPs. ZP pellets were pooled and centrifuged to obtain a single pellet. Isolated ZPs in a final volume of 100 -200 l were solubilized by adding ϳ2 l of 1 N HCl to lower the pH to 2.5 and incubating at 37°C for 15 min. Insoluble material was removed by centrifugation (16,000 ϫ g, 2 min). Solubilized ZPs were drop frozen and stored in liquid N 2 until use. To minimize protein loss caused by adsorption to surfaces, handling of ZPs, during isolation and in all subsequent experimental protocols, was performed using siliconized Pasteur pipettes, pipette tips, and microcentrifuge tubes.

I-Labeling of ZPs
Solubilized ZPs (approximately 80 g per labeling) were concentrated to 200 l in a vacuum concentrator (Labconco, St. Louis, MO), and the pH was adjusted to 7.0. ZPs were radioiodinated using Chloramine-T (Florman and Wassarman, 1985;Bleil and Wassarman, 1986). The free 125 I was removed, and the sample was desalted by size exclusion chromatography using an 8-cm Bio-Gel P6 Pasteur pipette column. Fractions containing 125 I-ZPs were pooled. The protein concentration of 125 I-labeled solubilized ZPs was determined by comparison to a protein standard curve (BSA) using the Quantigold protein assay (Diversified Biotech, Piscataway, NJ). Dilutions of 125 I-ZPs were quantified in a gamma counter to determine the counts/min per volume. The specific activity of the 125 I-ZP preparations ranged from 4.5 ϫ 10 5 cpm/ng to 1.8 ϫ 10 6 cpm/ng. Free 125 I contamination was estimated as a fraction of the total by densitometry of gel autoradiographs of ZPs and was less than 10% in all cases. Densitometry of the gel autoradiographs was also used to determine the contribution of 125 I-ZP3 to total radioactivity present in the labeled ZPs. Several exposures were made for each 125 I-ZP preparation, and only exposures within the linear range of the film (Kodak XRP-5) were used to calculate the fractional contribution of 125 I-ZP3. 125 I-ZP3 ranged from 16.3 to 22.4% of the total, and the precise contribution was calculated for every preparation of 125 I-ZPs used.

Sperm Collection
Epididymal sperm were collected from 12-15-week-old mice (Harlan Sprague-Dawley, San Diego, CA) by removing the epididymides, mincing the tissue in Hepes supplemented (20 mM, pH 7.2) (Florman and Storey, 1982); Whittingham's Buffer (HWB) (Whittingham, 1971), and gently agitating the suspension for 10 min to release sperm. Epididymal tissue was removed by centrifugation (200 ϫ g, 1 min). Sperm were concentrated by centrifugation (15 min, 200 ϫ g), resuspended in HWB containing 30 mg/ml BSA, and incubated at 37°C in a humidified atmosphere containing 5% CO 2 for 60 min to capacitate sperm. Sperm were collected after capacitation by pelleting (200 ϫ g, 10 min). Capacitated sperm were washed once by centrifugation in HWB without BSA, resuspended, and used in binding assays immediately or fixed. Sperm were fixed in 1% glutaraldehyde in HWB for 5 min and pelleted (250 ϫ g, 10 min). Fixed sperm were washed by pelleting and resuspension (three times) in fresh HWB, and finally resuspended in 500 l of HWB. Percent transmittance (%T) at 500 nm was determined for duplicate dilutions of the fixed sperm suspension. Average %T was used to calculate the sperm concentration from a standard curve of %T versus cells/ml (determined by hemacytometer counts of fixed sperm). Washed, fixed sperm were held at 4°C until use.

Binding Assays
Sperm and 125 I-ZPs were equilibrated to 37°C prior to experimental manipulations. Washed, fixed sperm were added to 125 I-ZPs and incubated in a water bath at 37°C. Sperm with bound 125 I-ZPs were separated from free 125 I-ZPs by vacuum filtration (Millipore, Bedford, MA) through Whatman GF/C glass fiber filters (nominal pore size, 1 m) that had been incubated in BSA (50 mg/ml in HWB, 1 h) to block nonspecific binding sites. Filters were washed with approximately 200 volumes (10 ml, with the exception of the off-rate experiments, as described below) of ice-cold HWB immediately. Filters (bound 125 I-ZPs) and samples of the filtrate (free 125 I-ZPs) were quantified using a Beckman 5000 gamma counter (Beckman Instruments). For all assays, parallel samples without sperm were processed identically, and the nonspecific 125 I-ZP binding to filters were quantified and subtracted from sperm-containing samples. All experiments were performed with duplicate samples. Following initial studies, 2 ϫ 10 6 sperm/sample was assigned to be the standard sperm aliquot for binding assays, since this number of cells reproducibly gives stable readings above background. Exceptions were experiments shown in Fig. 1, as noted below. Assay volumes varied from 50 to 200 l, but were identical for any given set of assays with the exception of the saturation assays, as noted below. Specific procedural details for each of the binding studies are outlined below.
Receptor Concentration Dependence-Increasing concentrations of live or fixed sperm were incubated for 15 min with a constant, low concentration of 125 I-ZPs.
IC 50 -Increasing amounts of unlabeled ZPs were added to saturating concentrations of 125 I-ZPs and incubated with sperm for 60 min. Samples to determine background binding for each concentration of unlabeled ZPs were processed in parallel.
Equilibrium Binding-Fixed sperm were incubated with 125 I-ZPs for various times ranging from 1-120 min.
On Rate-Aliquots of fixed sperm and 125 I-ZPs were incubated for increasing amounts of time from 1-10 min.
Off Rate-Fixed sperm were incubated with 125 I-ZPs for 60 min to attain equilibrium. Samples were pipetted into 10 ml of HWB in siliconized 15 ml plastic culture tubes and agitated on a Lab Line threedimensional Rotator for 10 -480 min at 37°C. The samples were vacuum filtered and washed with 5 mls of ice-cold HWB. Control samples (no dilution) were processed at the end of the equilibrium binding.
Saturation-Sperm were incubated with increasing amounts of 125 I-ZPs for 60 min. Assay volumes ranged from 60 to 300 l since the stock 125 I-ZP concentration was not great enough to maintain all the assay volumes at 60 l for the higher ZP concentrations.
Ca 2ϩ Dependence-Fixed sperm were aliquoted for equilibrium binding assays, pelleted in a standard microcentrifuge, and resuspended in HWB containing 1.7 mM Ca 2ϩ , or 2 mM EGTA with no added Ca 2ϩ ([Ca 2ϩ ] Ͻ 10 Ϫ7 M using a Ca 2ϩ electrode (Orion, Inc., Boston, MA). Aliquots of 125 I-ZPs were added for 60 min and samples separated by vacuum filtration and quantified. The composition of the wash buffers were identical to the incubation buffer for each sample (i.e. 1.7 mM Ca 2ϩ or 2 mM EGTA). Alternatively, sperm were collected and handled throughout in either a Ca 2ϩ containing HWB, or HWB with no added Ca 2ϩ and 2 mM EGTA. Since no significant differences (Student's t test, p Ͼ 0.05) in 125 I-ZP binding were detected.

Nonspecific Binding
GF/C filters were preincubated in various protein solutions to block nonspecific binding to filters and compared by vacuum filtration of 125 I-ZPs. The fraction (bound/free) of 125 I-ZPs retained following different protein blocks was determined (Table I) and, subsequently, all filters were preincubated with 50 mg/ml BSA for 1 h to block nonspecific binding sites on the filters. Nonspecific binding to BSA blocked GF/C filters by 125 I-ZPs or free 125 I was determined by vacuum filtration of 125 I-ZPs in the absence of sperm. This background binding was subtracted from experimental samples prior to data analysis. The contribution of free 125 I contamination to nonspecific binding to fixed sperm was assessed. 125 I-ZPs were separated from free 125 I using a Bio-Gel P6 column, as described above. Fractions from the 125 I-ZP peak and fractions from the free 125 I peak were incubated with 10 7 sperm for 1 h at 37°C and bound radioactivity determined as above. Less than 0.5% of the total radioactivity was associated with sperm under these conditions (data not shown).

Data Analysis
Experiments were replicated a minimum of three times and all replicates were performed in duplicate. Exceptions were the saturation experiment for which four independent data sets are pooled and the data points appearing on the plot are binned data averages presented with error bars (ϮS.E.) for both x and y components. Curve fitting was performed using TableCurve2D (Jandel Scientific, San Rafael, CA) and plotted with SigmaPlot (Jandel Scientific). Goodness of fit for all curve fitting, as indicated by the r 2 values, is stated in the figure legends. All calculated affinity parameters (B max , EC 50 , k on , and consequently K d1 and K d2 ), used the assumption that ZP3 is the only component binding to the intact mouse sperm, as indicated by previous studies (Bleil and Wassarman 1980a, 1980bFlorman et al., 1984;Florman and Wassarman, 1985;Mortillo and Wassarman, 1991). Concentration dependence data ( at which 50% of the 125 I-ZPs have been displaced. The K I was calculated using the correction of Cheng and Prusoff (1973), so that K I ϭ IC 50 /(1 ϩ ([L]/EC 50 )). Equilibrium binding data (Fig. 3) was fit to the square hyperbola: On rate data (Fig. 4) was fit by linear regression with k on ϭ ⌬B/ is the quantity of ligand present per assay and R T is the total number of receptors present, calculated using the value of 30,000 R/sperm (see "Results"). The y intercept (extrapolated value of bound counts/min at t ϭ 0) for these data was assumed to be the nonspecific background binding to sperm and was subtracted from all data points for the plot presented here. Off rate data (Fig. 5) was fit to a double exponential decay equation B/B 0 ϭ B 1 e Ϫk1t ϩ B 2 e Ϫk2t , where B is the amount bound at time t after dilution, B 0 is the amount bound prior to dilution, B 1 and B 2 are the fractions of receptor in high and low affinity sites, and k 1 and k 2 the off rate constants for each. The saturation binding isotherm ( Fig. 6) was fit to B ϭ (B max )[L] n /((EC 50 ) n ϩ [L] n ) where n is the Hill coefficient (Hulme and Birdsall, 1992). The raw binding data was transformed to units of "molecules of ZP3 bound" using the published molecular weight of 83 kDa for ZP3 and assuming that all bound radioactivity was due to ZP3, since extensive studies have indicated that only ZP3 binds to intact sperm (Bleil and Wassarman, 1980a, 1980bStorey, 1982, Florman et al., 1984;Wassarman, 1985, Mortillo andWassarman, 1991). The concentration of free ZP3 was derived using the percent contribution of 125 I-ZP3 to the total radioactivity of the sample, determined by gel autoradiography as discussed above. Additionally, saturation data was transformed to log (B/B max Ϫ B) and log (free[ZP3]) and plotted for determination of the Hill coefficient. The data were fit by linear regression for each phase and the slopes are the Hill coefficients. Significant differences in binding in the presence or absence of Ca 2ϩ were determined by an unpaired Student's t test using a significance level of p Ͻ 0.05.

RESULTS
Quantitative characterization of sperm-zona pellucida binding has been lacking although the zona ligand, ZP3, was identified and initially characterized over a decade ago Wassarman, 1980a, 1980b;Florman and Wassarman, 1985). The paucity of quantitative information concerning receptorligand interactions during fertilization stems from two intrinsic difficulties. First, although methods have been developed for isolating zonae pellucidae, only minute amounts are available for detailed binding studies. A mouse will typically yield less than two hundred zonae containing approximately 400 ng of ZP glycoproteins. 2 Second, in addition to its role as an adhesion molecule, ZP3 also functions as an agonist (secretagogue) for acrosomal exocytosis. The exocytotic event may lead to a loss (or modification) of receptor binding sites following exocytosis. Since this event occurs over the first 10 -30 min, finding conditions which are optimal for binding studies (e.g. equilibration and saturation) is nearly impossible. Because of these inherent problems, we have developed an assay using glutaraldehyde-fixed acrosome-intact sperm to quantitatively establish the number of ZP binding sites on sperm and the affinity of the ligand for its complementary receptor.
Before carrying out binding assays, conditions for minimizing nonspecific binding of ZPs to the glass-fiber filters needed to be established. Filters were incubated in protein solutions at different concentrations in order to block potential nonspecific binding interactions with solubilized ZPs. Blocking with BSA or casein resulted in low background binding to the filters (Table I). In contrast, ␣-lactalbumin and ovalbumin-blocked filters retained significantly more 125 I-ZPs compared to 50 mg/ml BSA (p Ͻ 0.01) and did not appear to be suitable blocking agents. Although casein blocking resulted in slightly lower background binding, the retention of insoluble aggregates of this protein by the filter suggested that it might not be a suitable choice. We therefore used 50 mg/ml BSA for blocking GF/C filters in all experiments.
To demonstrate that glutaraldehyde fixation of sperm did not affect ZP binding, different concentrations of both fixed and unfixed sperm were incubated with solubilized, 125 I-labeled ZPs for 15 min. The low concentration of ZPs along with relatively short incubation times were chosen so that acrosomal exocytosis would not occur during this experiment. Indeed, using higher concentrations of ZPs, the temporal binding pattern was highly complex and did not reach equilibrium, most 2 Q. Chen and R. A. Cardullo, unpublished data. likely reflecting changes in sites as sperm initiated acrosomal exocytosis (Thaler and Cardullo, 1994). However, at lower ZP concentrations and shorter times, both fixed and unfixed sperm displayed a linear dependence of ZP binding on sperm concentration, i.e. receptor number (Fig. 1). In addition, unlabeled ZPs could compete with the 125 I-ZPs for binding sites on fixed sperm (Fig. 2), indicating that binding specificity is retained by 125 I-labeled ZPs and that both labeled and unlabeled ZPs compete for the same site on the fixed sperm models. These data demonstrate that the fixed sperm model is useful for quantifying sperm-zona interactions prior to acrosomal exocytosis. Further, a K I for unlabeled ZPs was calculated from the IC 50 , the concentration of unlabeled ZPs at which 50% of the 125 I-ZPs are displaced from their binding sites under equilibrium conditions. The K I for ZP binding was found to be 3.78 nM by the method of Cheng and Prusoff (1973) and is in close agreement with the EC 50 of 1.29 nM, determined by steady state saturation (see Fig. 6). In contrast to untreated sperm, glutaraldehyde-fixed sperm achieved equilibrium binding, as evidenced by a plateau in solubilized 125 I-ZP binding during longer incubations (Fig. 3). In this representative experiment, the halftime to equilibrium was 22.1 min, which represents an upper limit for the time to reach equilibrium since the ZP concentration was below saturation.
The forward rate constant (k on ) for ZP binding was deter-mined by incubating sperm with 125 I-ZPs for various lengths of time, measuring the 125 I-ZPs bound and calculating the initial rate of binding (Fig. 4). Assuming that all specifically bound radioactivity is due only to ZP3, that the rate of binding is a first-order reaction (v on ϭ k on [ZP3][ZP3R]), and that the reverse reaction is negligible at early time points, the calculated on-rate constant was 3.20 ϫ 10 6 M Ϫ1 min Ϫ1 at 37°C. Dissociation experiments were used to determine the reverse rate constant, k off . Fixed sperm were incubated with 125 I-ZPs for 60 min to reach equilibrium binding, at which point sperm were diluted into a large volume of buffer and the decrease in 125 I-ZPs bound over time was quantified (Fig. 5). Loss of radiolabeled ligand was clearly biphasic and followed an exponential decay function with high fidelity. Rate constants for the low and high affinity components of 125 I-ZP binding were calculated to be k off1 ϭ 0.161 min Ϫ1 and k off2 ϭ 0.0023 min Ϫ1 , respectively. These rates correspond to 1 ⁄2 values for dissociation of 4.3 min and 300.8 min for the low and high affinity components. Steady-state saturation experiments using increasing concentrations of 125 I-ZPs incubated with a fixed number of sperm (2.0 ϫ 10 6 sperm) were performed under equilibrium binding conditions to determine the total number of binding sites (B max ), the presence or absence of cooperative binding effects, and the overall affinity of ZP binding (EC 50 ) for its complementary receptor on the mouse sperm surface. The amount of  Cheng and Prusoff (1973), was found to be 3.78 nM. Residual binding of 0.17 could not be displaced and was considered nonspecific binding. These data are consistent with the assumption that both the 125 I-ZPs and the unlabeled ZPs recognize the same binding site and further demonstrate that fixed sperm retain ZP binding sites.

FIG. 3. Equilibrium binding of 125 I-ZPs to fixed mouse sperm.
Solubilized 125 I-ZPs reached equilibrium binding with fixed spermatozoa with a half time of approximately 25 min at 37°C. This data set shows results of one representative determination with a half-time of 22.1 min. The mean 1 ⁄2 calculated from three independent data sets was 24.5 Ϯ 2.3 min.

TABLE I Blocking nonspecific binding of 125 I-ZPs to GF/C filters
Solubilized 125 I-ZPs were vacuum filtered through GF/C filters blocked for 1 h at room temperature with the indicated concentrations of protein. Blocking with either BSA or casein resulted in low background binding to filters. In contrast, ␣-lactalbumin and ovalbumin incubated filters retained significantly more 125 I-ZPs, compared to 50 mg/ml BSA (PϽ0.01) and did not appear to be suitable blocking agents. Although casein blocking resulted in slightly lower background binding, the retention of insoluble aggregates of this protein by the filter suggested it might not be the most suitable choice. We therefore used 50 mg/ml BSA for filter blocking in all experiments. 125 I-ZP bound to the fixed sperm surface increased with increasing free ZP concentration until available binding sites were saturated (Fig. 6A). These studies represent the first direct quantitative determination of the number of ZP binding sites on the mouse sperm surface and, in addition, concur with the kinetic studies in revealing that binding is more complex than can be suitably fit to a single-site binding model. The data were fitted to a nonlinear isotherm equation of the form B ϭ B max [L] n /((EC 50 ) n ϩ [L] n ) (Hulme and Birdsall, 1992), where n is the Hill coefficient, B is assumed to be ZP3 bound, and [L] is the concentration of free ZP3 present in the assays. The data fit well (r 2 ϭ 0.92) to this multisite model with n ϭ 1.72, a B max of 30,000, and an EC 50 of 1.29 nM. Sigmoidal binding curves, such as the one shown in Fig. 6A, are indicative of complex binding interactions between receptors and ligands. Accordingly, other measures of binding complexity were analyzed. Fractional occupancy rates, Scatchard plots, and Hill plots are traditionally viewed as indicators of cooperativity, although they can also reflect other complex mechanisms that yield similar plots and are usually lumped into the alternative category of "apparent cooperativity." Fractional occupancy of the sperm-ZP3 binding sites increased from 10 to 90% within less than 1.9 log units of free ligand (1.5 log units for sperm-ZP3), suggesting cooperative, or multisite, interactions (Wells, 1992). Scatchard plots displayed a concave upward shape (data not shown), an indicator of multiple site interactions as a result of the presence of multivalent ligand and/or receptor (Lauffenburger and Linderman, 1993). Hill plots of the equilibrium saturation data produced a biphasic linear curve (Fig. 6B), also supporting multivalent interactions (Matthews, 1993). Together with the kinetic data, these steady state saturation experiments support a complex binding interaction between ZP3 and its complementary receptor on the sperm surface.
Other researchers have demonstrated that the sperm binding bioactivity resides solely with ZP3 (Bleil and Wassarman 1980a, 1980bFlorman et al., 1984;Florman and Wassarman, 1985;Mortillo and Wassarman, 1991), and we have confirmed this using a chemical cross-linker. 3 When the total 125 I signal is attributed to ZP3, we calculated a B max of 30,000 ZP3 binding sites per mouse sperm (Fig. 6). In each experiment, the fractional radioactivity assigned to each of the three glycoproteins was determined. On average, ZP3 accounted for approximately 19% of the total radioactivity (range ϭ 16.3 to 22.4%) and the ZP3 contribution was used to determine the relative concentration of free ZP3 in the assay samples. The half-maximal concentration of ZP3 needed to achieve saturation, EC 50 , was found to be 1.29 nM. The K d values for high and low affinity binding calculated from kinetic data are 0.72 nM and 50 nM, respectively, demonstrating the existence of multiple binding sites (Table II). Taken together, results from both the kinetic and steady state experiments demonstrate that the initial adhesion event between the sperm and the zona pellucida is a complex, high affinity binding interaction.
Finally, equilibrium binding studies characterizing the Ca 2ϩ dependence of sperm 125 I-ZP binding were conducted and no changes in binding were detected in the presence or absence of Ca 2ϩ (data not shown). Previous work has suggested that Ca 2ϩ is required for binding of live sperm to intact zonae (Saling et al., 1978) but these earlier qualitative experiments could not distinguish between primary adhesion events involving acrosome-intact sperm and secondary binding events following acrosomal exocytosis. We were not able to detect a Ca 2ϩ requirement for binding of intact sperm to zona pellucida glycoproteins using the fixed cell models, but the applicability of this result to live cells and any requirements for Ca 2ϩ at subsequent steps during sperm-zona interactions remains to be determined.
These studies demonstrate that fixed sperm retain specific ZP binding sites and have enabled us to characterize the initial sperm-ZP binding interactions. Our data indicate that the interaction has both high and low affinity components and appears to be complex in nature. In addition, this binding assay establishes a standard quantitative method to evaluate putative sperm ZP3 receptors. Ligand dissociation experiments were performed to determine the off rate constant(s) for 125 I-ZP binding. Following equilibrium binding of fixed sperm and 125 I-ZPs, samples were diluted into a large volume of buffer and the loss of bound radiolabeled ligand (B/B 0 ) over time was quantified. Ligand dissociation fit a double exponential decay function from which off rate constants, k off1 and k off2 , for low and high affinity binding sites were determined to be 0.161 min Ϫ1 and 0.0023 min Ϫ1 , respectively (r 2 ϭ 0.98). These rates correspond to half-times of dissociation of 4.3 min for the low affinity and 300.8 min for the high affinity components of sperm ZP binding. The data points are averages of three independent experiments, each performed in duplicate, Ϯ S.E.

DISCUSSION
Previous descriptions of sperm-zona adhesion have relied on microscopic binding assays using living sperm and intact zonae pellucidae. In these assays, sperm are typically incubated with eggs for some predetermined amount of time (e.g. 5-60 min), and attached sperm are subsequently counted under the light microscope. This assay successfully identified mZP3, an 83-kDa glycoprotein from the mouse zona pellucida, as both the initial adhesion ligand and as a secretagogue for acrosomal exocytosis in mice (Bleil and Wassarman, 1980b: Florman et al., 1984, Florman and Wassarman, 1985. Using solubilized ZPs, a number of researchers have shown that ZP3 is the only component that binds to acrosome intact sperm (Bleil and Wassarman, 1986;Mortillo and Wassarman, 1991) and that concentrations of ZP3 exceeding 1 ng/l (equivalent to ϳ1 zonae/l) are sufficient to block sperm-FIG. 6. Steady state saturation binding of 125 I-ZPs to fixed mouse sperm. A, equilibrium binding saturation experiments were conducted to independently determine the binding affinity and complexity, as well as the B max for sperm-ZP binding. Free ZP3 concentration was calculated from ZP3 fractional contribution to total 125 I-ZPs. Bound radioactivity was assumed to be ZP3 and conversion to numbers of ZP3 molecules made accordingly (see "Experimental Procedures" for details). The saturation binding isotherm demonstrates that there are approximately 30,000 ZP3 binding sites per sperm and that the ZP3 concentration required for 50% receptor occupancy, EC 50 , is 1.29 nM. The Hill coefficient for the binding isotherm is 1.72. r 2 ϭ 0.92. The data are binned data points from four independent experiments, plotted Ϯ S.E. for both bound and free ZP3. B, a Hill plot generated from the saturation data revealed that the binding of solubilized ZPs was more complex than could be explained by a simple bimolecular interaction between a univalent ligand and its complementary receptor on the sperm surface. Nonlinear Hill plots can indicate multivalent binding events or cooperativity. This conclusion was reached by Scatchard analysis as well (data not shown). Best fits for each portion of the curve are n ϭ 0.6 (r 2 ϭ 0.88) and n ϭ 2.3 (r 2 ϭ 0.89). zona binding as well as initiate the signal transduction pathway leading to acrosomal exocytosis (Bleil and Wassarman 1980a, 1980bFlorman et al., 1984;Florman and Wassarman, 1985;Leyton and Saling, 1989a).
Although useful for studying the native interactions between mZP3 and its complementary receptor on the sperm surface, using intact zonae pellucidae may preclude accurate determination of binding characteristics between ligand and receptor. The quantitative, standardized binding assay presented here, using solubilized ZP glycoproteins, has revealed molecular details about the specific interactions between mammalian sperm and egg and will aid in the identification of bona fide ZP3 receptors on acrosome-intact sperm.
Initial attempts to characterize solubilized ZP binding to living mouse sperm proved difficult because equilibrium was never achieved (Thaler and Cardullo, 1994). This was presumably due to loss of binding sites following the acrosome reaction (Bleil and Wassarman, 1986;Mortillo and Wassarman, 1991). Removal of binding sites is common in living cells and, in such cases, binding events can be distinguished from downstream physiological processes by blocking the removal of binding sites through use of low temperatures, pharmacological inhibitors, or fixation. In our studies, ZP binding experiments below 15°C were precluded since solubilized ZPs precipitated out of solution under these conditions (data not shown). Pharmacological inhibitors, including pertussis toxin (Endo et al., 1987(Endo et al., , 1988) and 3-quinuclidinyl benzilate (Florman and Storey, 1981) block ZP3 induced acrosomal exocytosis, but the rate of spontaneous acrosomal exocytosis is unaffected and is sufficient to significantly decrease the fraction of acrosome-intact sperm over reasonable binding times (Florman and Storey, 1982). In contrast, fixed cell models have been widely used to characterize ligand receptor interactions including those of growth factor (van Maurik et al., 1985;Zidovetzki et al., 1991), cytokine (Smith et al., 1979), adrener-gic (von Zastrow andKobilla, 1992), and acetylcholine (Schroder et al., 1990) receptors. In this study, we have shown that 125 I-labeled ZPs bind to glutaraldehyde-fixed sperm in a concentration dependent manner, that unlabeled ZPs compete for binding sites of 125 I-ZPs on mouse sperm, and that binding of 125 I-ZPs at low concentrations and short times (conditions that do not favor ligand-induced acrosomal exocytosis) are virtually identical for live and fixed sperm. These experiments demonstrate that glutaraldehyde-fixed mouse spermatozoa are a good model system for characterizing the initial binding events between sperm and zona pellucida and provide a stable population of acrosome intact sperm (typically greater than 80%) for binding experiments.
Both kinetic and equilibrium binding studies undertaken to quantify the sperm-ZP binding interactions were performed using whole solubilized 125 I-labeled ZPs containing all three ZP glycoproteins. The binding parameters calculated from these studies have assumed that all specifically bound radioactivity is solely due to the binding of ZP3 as suggested by previous work (Bleil and Wassarman 1980a, 1980bFlorman et al., 1984;Florman and Wassarman, 1985;Mortillo and Wassarman, 1991). Studies quantitatively determining the binding parameters of individual ZP glycoproteins are planned pending availability of cloned ZPs since such studies require great quantities of ZP glycoproteins and are not feasible with biochemically isolated material.
Kinetic studies to determine the forward and reverse rate constants for sperm-ZP binding suggest a complex interaction for this receptor-ligand pair. The slow association kinetics indicated by the k on of 3.2 ϫ 10 6 M Ϫ1 min Ϫ1 could indicate a complex mechanism requiring the interaction of multiple ligands and/or receptor subunits, or changes in conformational state during binding, as has been suggested for other ligandreceptor interactions, such as transferrin (Ciechanover et al., 1983), insulin (Lipkin et al., 1986), and IgE (Pruzansky and Patterson, 1986), in which the forward rate constant is significantly below the diffusion limited rate for surface bound receptors and soluble ligands (Lauffenburger and Linderman, 1993). Dissociation experiments delineating the off rate constants demonstrate the presence of high and low affinity components of sperm-ZP binding, again pointing to the complexity of this interaction. Analysis of equilibrium saturation studies suggests a multisite mechanism for binding, and Scatchard and Hill analysis suggest possible cooperative interactions. It is not possible, however, to differentiate between true cooperativity (i.e. a change in affinity dependent upon the occupancy state of the receptor population) or apparent cooperativity (due to other complex mechanisms including multivalent receptors or multivalent ligands) from these data alone. However, other research has strongly indicated that ZP3 is a multivalent ligand, since proteolytically derived, monovalent glycopeptides of ZP3 bind to sperm but are not able to induce acrosomal exocytosis (Florman and Wassarman, 1985), but subsequent antibody cross-linking of ZP3 glycopeptides results in acrosomal exocytosis (Leyton and Saling, 1989a). Additionally, monovalent oligosaccharides do not inhibit sperm-ZP binding, but synthetic multivalent oligosaccharides show an enhanced ability to block sperm binding to intact ZPs as the number and length of branches is increased (Litscher et al., 1995). These data may suggest that the binding behavior observed here in both kinetic and steady state experiments is due to an apparent cooperativity resulting from the multivalent nature of the ligand for this system.
Evidence which may suggest the presence of multiple receptors in addition to a multivalent ligand comes from the observations that sperm-ZP adhesion and, consequently, fertiliza- Kinetic 3.20 ϫ 10 6 0.161 0.0023 50 0.72 tion can be blocked using substrates and antagonists against a number of sperm surface molecules including ␤-1,4-galactosyltransferase Hall, 1982a, 1982b;Shur and Neely, 1988), mannosidase (Cornwall et al., 1991;Tulsiani, et al., 1989), fucosyltransferase , and the trypsin inhibitor-sensitive site (Benau and Storey, 1987). While it is widely recognized that the binding activity of ZP3 resides within O-linked oligosaccharides of ZP3 (Bleil and Wassarman, 1980b;Florman and Wassarman, 1985;Litscher and Wassarman, 1993) and not with the highly conserved polypeptide chain (Chamberlin and Dean, 1990;Litscher and Wassarman, 1996), the specific saccharide residues responsible for sperm-ZP binding remain controversial. Some evidence suggests that the terminal monosaccharide directly responsible for adhesion is an ␣-galactose Shalgi et al., 1991;Litscher et al., 1995), but transgenic mice lacking all 1,3-␣galactose epitopes were fertile (Thall, et al., 1995), suggesting that ␣-galactose may not be solely responsible for the sperm-ZP interactions required for fertilization. Other evidence suggests that the terminal monosaccharide is a ␤-GlcNAc which acts as an acceptor for the sperm surface ␤-1,4-galactosyltransferase (Shur and Hall, 1982;Shur and Neely, 1988;Miller et al., 1992), but synthetic oligosaccharides containing GlcNAc in a ␤-linkage at the nonreducing terminus had no effect on spermzona binding (Litscher et al., 1995). Despite the obvious contradictions among these studies, it is clear that none of these oligosaccharide blockers effectively inhibit sperm binding to intact ZPs, as the IC 50 values reported are in the micromolar to millimolar range. One hypothesis arising from such studies is that multiple adhesion molecules are involved in sperm-zona binding, and the high IC 50 values reflect a number of low affinity interactions. The complex nature of the sperm-ZP binding isotherm presented here may thus reflect the involvement of one or more of these receptors leading to the formation of a fertilization complex that is required for adhesion and/or acrosomal exocytosis. Alternatively, the high IC 50 values derived from these assays may be due to limitations intrinsic to quantifying a competitive interaction between a ligand embedded within a matrix (the intact ZP) and a soluble competitor. Further, live, acrosomeintact, sperm and zonae are never in equilibrium above a critical ZP3 concentration because acrosome reactions are initiated rapidly and determining an IC 50 or K I for agents interfering with binding is therefore impossible. Consequently, an assay using fixed sperm, or some method which prevents the acrosome reaction but not ZP binding in live sperm, is absolutely required to identify and characterize putative receptors. The K I determined for unlabeled ZP displacement of 125 I-ZPs is in close agreement with the EC 50 from saturation binding studies and this suggests that the soluble binding assay is also a suitable approach to determine the relative affinities of ZP3 and its putative competitors for binding sites on the sperm surface. The conditions and binding assay presented in this study should provide guidelines for testing models of sperm-ZP binding in a quantitative fashion.
While the results presented here are essential for characterizing sperm-zona interactions at a molecular level, they may only partially address the interactions that occur on the intact zona pellucida. Theoretical arguments based on biophysical measurements of sperm adhesion have suggested that only a few sperm-zona bonds are needed to tether a sperm to the egg (Baltz et al., 1988). The rate-limiting step for sperm adhesion on the intact zona pellucida is related to the surface density of ZP3 on the intact zona, the concentration of complementary ZP3 receptors on the sperm surface, the contact area between sperm and zona and the diffusion coefficient of the membrane bound receptor on the sperm surface (Baltz and Cardullo, 1989). The ZP3 density has been calculated to be as high as 300 molecules/m 2 and the contact area has been calculated to be from 0.1 to 5 m 2 (Baltz and Cardullo, 1989). However, no number has been determined for either the ZP3 receptor density on the sperm surface or its diffusion coefficient. The data presented here demonstrate that the total number of ZP3 binding sites is approximately 30,000 per sperm, and we are now using a recently developed fluorescent conjugate of ZP3 (Chen and Cardullo, 1994) to determine both the density of ZP3 binding sites on the sperm surface along with the mobility of these molecules in order to more precisely examine the initial interactions between sperm and egg. At present, the high concentration of ZP3 presented by the intact ZP combined with the relatively high number of ZP3 binding sites on sperm predicts that adhesion between sperm and egg is inevitable once physical contact is made. In concert with receptor localization, continued studies in our laboratory are underway to evaluate the mechanism of sperm-ZP interactions and to identify the mouse sperm receptor for ZP3 using the soluble binding assays described here.
In summary, the binding data presented here show, by several analyses, that the nature of sperm binding to the zona pellucida is complex, resulting in the apparent cooperativity observed in sperm-ZP binding. The complex nature of the sperm-ZP binding isotherm may thus reflect the interaction of multiple sperm proteins with a multivalent ZP3, leading to the formation of a fertilization complex that is required for adhesion and/or acrosomal exocytosis.